Skip to main content

Table 5 Absolute bias in arms dropped at the first interim analysis

From: A multi-arm multi-stage clinical trial design for binary outcomes with application to tuberculosis

   

I =D = culture status

I = culture status,D = relapse

α 1

Treatment

% stop

Trueθ D

Bias onD

Bias onDafter f/u

Trueθ D

Bias onDafter f/u

 

arm

at

 

at interim

Length of f/u

Length of f/u

 

Length of f/u

Length of f/u

  

stage 1

 

analysis

onI= 4 wks

onI= 8 wks*

 

onI= 4 wks

onI= 8 wks*

  1 power ω 1 = 90%

0.5

A

65

-5%

-6.9%

-4.7%

-3.2%

-10%

-1.4%

-1.0%

 

B

49

0%

-9.5%

-6.5%

-4.6%

-6%

-1.7%

-1.3%

 

C

23

8%

-14.6%

-9.8%

-7.0%

-3%

-2.7%

-1.9%

 

D

10

13%

-18.2%

-12.3%

-8.8%

0%

-3.0%

-2.1%

0.2

A

94

-5%

-0.9%

-0.8%

-0.7%

-10%

-0.2%

-0.2%

 

B

80

0%

-2.6%

-2.1%

-1.9%

-6%

-0.5%

-0.5%

 

C

35

8%

-6.9%

-5.9%

-5.0%

-3%

-1.6%

-1.4%

 

D

10

13%

-10.7%

-9.2%

-7.7%

0%

-2.2%

-2.0%

  Stage 1 power ω 1 = 95%

0.5

A

70

-5%

-4.6%

-3.6%

-2.8%

-10%

-1.0%

-0.9%

 

B

49

0%

-7.3%

-5.6%

-4.5%

-6%

-1.5%

-1.2%

 

C

17

8%

-12.6%

-9.8%

-7.6%

-3%

-2.7%

-2.1%

 

D

5

13%

-16.3%

-12.9%

-9.9%

0%

-3.4%

-2.7%

0.2

A

95

-5%

-0.6%

-0.6%

-0.5%

-10%

-0.2%

-0.2%

 

B

80

0%

-2.1%

-1.8%

-1.6%

-6%

-0.5%

-0.4%

 

C

27

8%

-6.7%

-6.0%

-5.3%

-3%

-1.7%

-1.5%

 

D

5

13%

-10.8%

-9.6%

-8.4%

0%

-2.3%

-2.1%

  1. Simulation results showing the proportion of trials stopped at the first interim analysis and the absolute bias for such arms in the estimated treatment effect on D at the interim analysis and after all remaining patients have been followed up. Key: α 1 = significance level in stage 1, θ D = underlying treatment effect on the definitive outcome. *Plus an additional 6 week delay to determine culture status.